In order to obtain direct evidence that endotoxin-induced tumor regression is based on T cell-mediated, concomitant antitumor immunity, it was shown, first, that the rise and fall of susceptibility of the SA-1 sarcoma to endotoxin treatments (50 microgram intravenously) corresponded exactly to the generation and decay of the host's concomitant immune response to this tumor and that complete regression occurs only when endotoxin is given at the peak of concomitant immunity on day 9 to 10. Secondly, it was shown that endotoxin fails to cause the regression of the SA-1 sarcoma growing in mice that are incapable of generating concomitant immunity because of having been made T-cell deficient (TXB mice) by thymectomy and lethal irradiation and restored with bone marrow. Most importantly, it also was shown that endotoxin causes the complete regression of large SA-1 tumors growing in TXB recipients provided these recipients are given spleen cells from immunocompetent 9-day tumor-bearing donors, i.e., from donors with peak concomitant immunity and, therefore, with susceptible tumors. The spleen cells from tumor-bearing donors that mediate endotoxin-induced tumor regression in TXB recipients are destroyed by monoclonal anti-Thy-1.2, or anti-Ly-2.2 antibody and complement, but not by anti-Ly-1.2 antibody and complement. This represents unequivocal direct evidence that endotoxin-induced tumor regression is based on a T cell-mediated immune response to tumor-associated transplantation antigens. Reciprocal passive transfer experiments with the SA1 and a benzpyrene-induced fibrosarcoma have revealed, in addition, that the mediation of endotoxin-induced regression is specific. It can be hypothesized, therefore, that endotoxin and other agents that are capable of causing the regression of already established immunogenic tumors are immunofacilitators rather than immunopotentiators in that they facilitate the expression of an already acquired state of concomitant immunity. Results obtained thus far with Poly I:C show that the antitumor properties of this compound likewise are based on its capacity to facilitate the expression of an acquired state of concomitant immunity. It can be predicted from this hypothesis that any manipulation that augments concomitant antitumor immunity will make a tumor more susceptible to the antitumor action of endotoxin and Poly I:C. (TA)